Output: What NEWGARDEN Tallies, Calculates, and Reports - Genetic Diversity in Establishing Plant Populations: Founder Number and Geometry - page 57

5

CHAPTER

Output: What NEWGARDEN
Tallies, Calculates, and
Reports

A “trial” is defi ned as the generation of one complete set of results based on
the information entered by the user in a NEWGARDEN input fi le. For each
trial, NEWGARDEN produces a table of summary statistics for analyses of
that trial in the form of an output fi le. The user can specify the number of
“runs” for a trial, each run being an independent replicate of population
development starting from a newly generated set of founders and growing
under the user-specifi ed conditions for the trial. For example, the user may
specify 25 replicate runs for a given set of trial conditions. After completing
this trial and collecting the output summary statistics based on those 25
replicate runs, the user can then change one of the initial trial conditions,
say, changing selfi ng rate from 0 to 25%, again specifying 25 replicate runs
of the new trial conditions. The output fi les for these two trials can then be
compared to examine the effects of this change in selfi ng rate since a mean
and standard deviation (s.d.) for output statistics are calculated separately
for each trial based on the results generated across the 25 replicate runs
for a given trial.
Setting a higher number of runs per trial (say, 100 runs in the example
just given) will give more accurate estimates of means with s.d. of reported
statistics, but also increases the computer time needed. The calculation
of means with associated s.d. allows the user to statistically compare the
effects of changes in a particular trial input condition (e.g., changing selfi ng
rates between trials) while holding other conditions of comparative trials
constant. Standard deviations may also be compared across trials to gain
some sense of comparable variability or risk of changes in different trial
input conditions. For example, if there is greater variation (as measured
by the s.d.) in the amount of genetic diversity preserved in a trial when
planting founders in a square than in an identical trial with the same
number of founders planted in three relatively long lines, then even if
there is no appreciable difference in the means of these two trials, there
may be more risk associated with planting in a square. In other words, in